NCRP report no 110 some aspects of strontium radiobiology

For isotopes that emit beta particles of long-range in tissue, such as 90Sr + it is probably sufficient to use the total residence times for either compact or cancellous bone, respectiv

NATIONAL COUNCIL O N RADIATION

PROTECTION AND MEASUREMENTS

Issued August 31, 1991

National Council on Radiation Protection and Measurements

LEGAL NOTICE This report was prepared by the National Council on Radiation Protection and Mea- surements (NCRP) The Council strives to provide accurate, complete and useful information in its reports However, neither the NCRP, the members of NCRP, other persons contributing to or assisting in the preparation of this report, nor any person acting on the behalf of any of these parties: (a) makes any warranty or representation, express or implied, with respect to the accuracy, completeness or usefulness of the information contained in this report, or that the use of any information, method or process disclosed in this report may not infringe on privately owned rights; or (b) assumes any liability with respect to the use of, or for damages resulting from the use

of any information, method or process disclosed in this report, under the Civil Rights

Act of 1964, Section 701 et seq as amended 42 U.S.C Section 2000e et seq (Title VII)

or any other statutory or common law theory governing liability

Library of Congress Cataloging-in-PubLication Data

National Council on Radiation Protection and Measurements

Some aspects of strontium radiobiology

p cm.-(NCRPreport ; no 110)

Prepared by Scientific Committee 57-12 on Radiostrontium

Includes bibliographical references and index

ISBN 0-929600-19-3

1 Strontium-Isotopes-Toxicology 2 Strontium in the body

3 Tumors, Radiation-induced 4 Bones-Cancer-Animal models

I National Council on Radiation Protection and Measurements

Scientific Committee 57-12 on Radiostrontium 11 Title

Preface

This report, which provides information on several aspects of stron- tium radiobiology, was prepared by Scientific Committee 57-12 In preparing the report, the committee was able to utilize unpublished material prepared some years ago by Scientific Committee 23 on Radiation Hazards Resulting from the Release of Radionuclides into the Environment Howard L Andrews, who served as a member of Scientific Committee 23, assisted Scientific Committee 57-12 during the preparation of this report and his help is gratefully acknowl- edged C.W Mays, now deceased, as well as others, also provided important information throughout

The report reviews pertinent information on the metabolism and dosimetry of radiostrontium Effects of radiostrontium are described briefly, especially as revealed in a series of long-term animal experi- ments The possible extrapolation of these results to man is consid- ered

Serving on Scientific committee 57-12 during the preparation of this report were:

Ray D Lloyd, Chairman

University of Utah Salt Lake City, Utah

Patricia W Durbin Roy R Pool

University of California University of California

Berkeley, California Davis, California

Robert K Jones Harvey A Ragan

Lovelace Biomedical and Battelle, Pacific Northwest

Environmental Research Institute Laboratory

Albuquerque, New Mexico Richland, Washington

John A Auxier,

Evaluation Research Corp

Oak Ridge, Tennessee

Lovelace Biomedical and

Environmental Research Institute

Albuquerque, New Mexico

Keith F Eckerman Roy C Thompson

Oak Ridge National Laboratory Battelle, Pacific Northwest

Richland, Washington NCRP Secretariat, E Ivan White

James A Spahn

The Council wishes to express its gratitude to the members of the Committees for the time and effort devoted to the preparation of this report

President, NCRP Bethesda, Maryland

March 31,1991

Contents

Preface

1 Purpose and Plan

2 Background

3 Metabolism of Radiostrontium

4 Dosimetry of 90Sr

5 Effects of =Sr as seen in the Long-Term Animal Studies

5.1 Studies with Mice a t Argonne National Laboratory

5.2 Beagles Injected a t the University of Utah

5.3 Beagles fed 90Sr a t the Laboratory for Energy Related Health Research (LEHR) University of California a t Davis

5.4 Beagles Exposed to Airborne 90Sr a t Inhaltion Toxicology Research Institute (ITRI)

5.5 Effects of 90Sr Fed to Miniature Swine a t Hanford (Battelle PNL)

5.6 Strontium-90 Exposure Via Injection in Monkeys (Univeristy of California Supplemental Data From the Univeristy of Rochester)

5.7 Summary and Discussion of Long-Term Animal Experiments with 90Sr

6 Effects of Y3r compared to =Ra

7 Genetic Effects of Y3r

8 Estimated Risks to Humans from Internally Deposited Y3r

8.1 Risk of Bone Cancer

8.2 Risk of Leukemia

8.3 Summary: Risk of Milignancy from Internally Deposited 90Sr

9 Conclusion

Appendix A Kilobecquerel-days Accumulated in Cancellous Bone in Cortical Bone, and in Soft Tissue Appendix B Retention Functions for wSr Retention in Skeltons

Appendix C Injection, Retention and Cause of Death for wSr Injected Monkeys University of California

vi 1 CONTENTS

Berkeley 60

References 64

The NCRP 77

NCRP Publications 84

Index 94

1 Purpose and Plan

This report provides: (a) a review of some aspects of the radiobiol- ogy of radiostrontium, especially strontium-90; (b) a review of the pertinent information on the metabolism and dosimetry of radio- strontium; (c) a brief description of the effects of radiostrontium, especially as revealed in a series of long-term animal experiments (on mice, dogs, swine, and monkeys) and their possible extrapolation

to man; and (dl estimates of risk to humans from the internal deposi- tion of 90Sr

While the information reviewed in this report is sufficient to sup- port first order estimates of risk to humans from internally deposited radiostrontium, the report does not recommend any changes in the radiation protection standards for strontium radionuclides This is partly because the current data do not indicate a need for major changes from current standards and partly because the reanalysis

of the risk factors for radiation derived from the revised Japanese exposure data may influence future radiation protection standards Thus, this report should be regarded as an important and necessary update but not a final interpretation for the purposes of standard setting

2 Background

Strontium-90 is a radioactive isotope of strontium that is produced

in nuclear fission with a relatively high yield of 3 to 4 percent (Glasstone and Dolan, 1977) It has a physical half-life of about 29 years and emits a beta particle of fairly low average energy (0.2 MeV) It is accompanied by a decay product, yttrium-90, which has

a shorter half-life (64h) and a much more energetic beta particle (2.3 MeV maximum, 0.93 MeV average) Because of its moderately long half-life, the energy of its radiation, its relatively high yield in the fission process, and its mobility under most circumstances, stron- tium-90 was early considered among the potentially most hazardous

of the products of nuclear fission Strontium-89 with a half-life of 50.5d and average beta energy of 0.58 MeV is also produced in nuclear fission along with several shorter-lived isotopes of strontium The possible hazard of 90Sr, with its relatively long physical half- life and its chemical similarity to Ca, was recognized in the work with atomic energy carried out as part of the Manhattan District efforts Experiments were undertaken very early to delineate the hazard more explicitly These were necessarily a t relatively high dose levels and of short term, because answers, even partial answers, were needed immediately Much of this work was the basis for the recommendations given in NCRP Report No 11 (NCRP, 1953) The interested reader is referred to that Report and to its bibliography

as well as to NCRP Report No 22 (NCRP, 1959), to ICRP Publication

No 2 (ICRP, 1959) and to the recently published history by Stannard (1988)

Initially, concern was for 90Sr in fallout from atmospheric tests of nuclear weapons By the mid-1970s concern had shifted to potential accidents with the 90Sr inventory in power reactors, at fuel reprocess- ing plants and in high-level waste, and the underlying concern for the potential consequences of nuclear weapons use The primary interest was in the induction of bone cancer by the beta radiations from radiostrontium, but the problems of fallout from nuclear weap- ons testing stimulated wide-ranging investigations of not only its effects but also its behavior in the environment, especially in food and food chains and in aqueous media and in a variety of organisms The radiobiology literature concerning the radioisotopes of stron- tium is enormous Nevertheless, for many years, little was known

2 BACKGROUND / 3

about its biological effects at doses low enough to be pertinent to the direct development of radiation protection standards Much analysis had to be done to establish relative effectiveness ratios in animals between strontium and radium whose long term effects in man at low doses had been well-established Some of this work was summa- rized in an unpublished report drafted by NCRP Scientific Commit- tee 23 which was used in preparing this report As these long term experiments developed, many progress reports were issued These include reports from symposia held at Sun Valley (Mays et al., 1969a1, a t Battelle Pacific Northwest Laboratories (Sikov and Mah- lum, 1969; Clarke et aZ., 1970a), at the University of California a t Davis (Goldman and Bustad, 1972) and at Glasgow and Strontian, Scotland (Lenihan, 1972) Also, as will become evident in the follow- ing text, the documents of the United Nations Scientific Committee

on the Effects of Atomic Radiation (UNSCEAR) and of the U.S National Academy of Sciences-National Research Council Commit- tee on the Biological Effects of Ionizing Radiation (BEIR) contain considerable information on the behavior and potential effects of radiostrontium Annual reports and occasional special documents from the several laboratories gave detailed reports in tabular form

of the status of each animal in each experiment

All of these sources provided useful information for the delibera- tions of governmental agencies and national and international bod- ies concerned with radiation protection Nevertheless, the slow pro- gression of the long-term animal experiments always lent a degree

of tentativeness to the conclusions that were possible It is against this background that the present report has been assembled While there are still some areas of uncertainty, it was considered useful to review the radiobiology of stontium without delay and to develop risk estimates for man

3 Metabolism of

Radiostrontium

Data on metabolism1 of radiostrontium (as well as stable stron- tium) have much intrinsic scientific interest and are, ofcourse, essen- tial for the calculation of radiation doses Since information on this aspect could be obtained much sooner than information on long-term effects, much of the earlier literature concentrated on metabolic behavior, but the long-term experiments also gathered information

on metabolism, particularly the kinetics of retention over long periods

The metabolic behavior of wSr in mammals can be described in general terms as follows: after radiostrontium is ingested, a fairly substantial part is absorbed from the gastrointestinal tract, and a part is excreted unabsorbed in the feces That which is absorbed is: (a) deposited in the bone volume; (b) distributed in an exchangeable pool which can be considered to be comprised of the plasma, extra- cellular fluid, soft-tissue and bone surfaces; or (c) removed from the body by urinary and fecal excretion Absorption of ingested stron- tium from the gastrointestinal (GI) tract by adult man averages between 20% and 30% (Spencer et al., 1960; Dolphin and Eve, 1963; Marshall et al., 1973; Likhtarev et al., 1975; Muth and Globel, 1983)

In biological systems, the behavior of stable strontium, and there fore of radioactive strontium that enters such systems, is qualita- tively similar to and is governed partially by the behavior of calcium Because of homeostatic control, there is a remarkable constancy of calcium concentration in most tissues and fluids (e.g., bone, blood and milk) Thus, it was proposed that within normal dietary ranges and under steady-state conditions, the radiostrontium-to-calcium ratio in the body tissues or secretions is a function of the ratio that exists in the diet (Comar and Wasserman, 1964; Comar, 1965,1967; Comar et al., 1955) For this reason, the concentration of radiostron- tium in biological materials was studied and reported by many inves- tigators as a SrtCa ratio (usually picocuries of wSr per gram of Ca)

l"Metabolism" means the behavior of the radionuclide in the organism; its absorp- tion, distribution, localization in cells and the tissues and its excretion

3 METABOLISM OF RADIOSTRONTIUM 1 5 This ratio was termed the "strontium unit" or "Sunshine unit" in the early years of work on fallout from nuclear weapons tests However, living organisms generally utilize and retain strontium less effectively than calcium (Shvedov, 1978) The term "discrimina- tion" has been used to describe the ratio of concentrations in, for example, the organism and its diet, and to denote the contributions

of individual physiological processes responsible for it The overall discrimination was designated as the "Strontium Calcium Observed

Ratio" (OR) (Comar et al 1956):

-

o k m p l e / p - W r - SrlCa Sr'ca of of precursor (dimensionless) (3.1)

This concept was used for many years and was a central feature

in predicting behavior of radiostrontium in organisms and in the environment (Comar and Wasserman, 1960; Stannard, 1988) How- ever, the behavior of strontium may also be considered directly with- out reference to calcium (Palmer et al., 1958a, 1958b; Palmer and

Thompson, 1961, 1964; Thompson and Palmer, 1960; Thompson, 1963)

To calculate the radiation dose to various parts of bone tissue from internally deposited radionuclides, one needs to know the pattern of uptake and retention of the radioactive material within the skeletal system Appropriate data for man are scarce, so that existing esti- mates of the dose within bone from the alkaline earth elements have been based on the skeletal distribution of fallout calcium distribution in the human skeleton and results of animal experi- ments One study, published by the ICRP (ICRP, 1972, also published

as Marshall et al., 1973), has considered the subject in detail and

used accumulated knowledge of the mechanisms of skeletal metabo- lism and the retention of radionuclides (Marshall, 1969) to construct

a quantitative metabolic model for the alkaline earth series It is reproduced in summary form in NCRP Report 84 (NCRP, 1985) This model (known as the ICRP or Marshall Model) deals with radionu- clides that distribute throughout the volume of mineralized bone, primarily the alkaline earths, calcium, strontium, barium and radium Obviously it cannot be applied to other chemical groups Data on the retention of both 90Sr and 226Ra in dogs from the experi- ments to be described presently a t Utah, a t the Laboratory for Health Related Research (LEHR), at the Inhalation Toxicology Research Institute (ITRI) and a t Argonne National Laboratory (ANL) were used in the construction of this model These data were generally described by exponential or modified exponential (power) retention functions (see below)

6 3 METABOLISM OF RADIOSTRONTIUM

The ICRP model has been extended to younger ages, modified or criticized by Papworth and Vennart (1973), Bennett and Harley (19731, Newton et al (1977), Marcus and Becker (1980), Matsubara etal., (1981), Harrison (1981), Schlenkeretal (1982), Johnson(1983), Holtzman et al (1983), Leggett et al (1982, 1984), and Crawford- Brown (1984), with additional published data by Likhtarev et al (1975), Wenger and Soukas (1975), Ilyin et al (1975), and Erre et al (1980)

Single dosage experiments in beagle dogs at ANL have given some indication that the young animal excretes less of its 'OSr than the adult It appears that a factor of 3 or 4 is sufficient to span the results (Decker et al., 1964, Finkel et al., 1972) This is in concurrence with the findings of the continuous feeding experiments at LEHR and PNL and with studies using injected 85Sr in beagles at Utah (Glad

et al., 1960) The influence of age on strontium metabolism (greater retention in younger individuals) has also been investigated by McClellan (1962), Speckman and Norris (19641, Della Rosa et al (1965), Forbes and Reina (1972), Kahn et al (1969), Papworth and Vennart (1973) and Leggett et al (1982, 1984)

Perhaps the most extensive data on the behavior of 'OSr in children, both in span of time and in the number of subjects studied, have been collected through the effort begun at the Health and Safety Laboratory (HASL) (now the Environmental Measurements Labora- tory), New York City, under the late Joseph Rivera, continued in subsequent years by Burton Bennett and developed into the Rivera Model (Rivera, 1969; Rivera and Harley, 1965) This is an empirical model based on measurements of fallout in the diet and in bone from individuals of different ages It is formulated as a simple description of the gain and loss of calcium and the associated Data of Mitchell (Mitchell et al., 1949) were used for the calcium content, and measurements of the concentrations of in bone and diet covering a considerable period of years were the basis for the 90Sr data in their study (Radiological Health Data and Reports, 1960- 1971)

A comprehensive review of information on the uptake and excre- tion of 'OSr by infants was published by Durbin et al (1970) Their analysis used data derived from the skeletal weight, Ca accretion and uptake, and the value of the observed ratio (OR) during the first year of life This was supplemented by values derived from Bedford

et al (19601, Mitchell et al (1949) and data of Beninson et al (1969)

as well as data from the HASL studies (e g HASL,1964) The dose commitment to red bone marrow was calculated by Durbin et al (1970) for the individual assumed to be always ingesting 37 kBq of 'OSr, and uptake from GI tract to blood was estimated by use of

3 METABOLISM OF RADIOSTRONTIUM 1 7

combined data of Durbin et al (1970) and Beninson et al (1969)

The dosimetric significance of some of the empirical models will be considered in Section 4

Qualitatively, the results of both the Rivera model and its revision, the Bennett model, are much the same, and they also agree qualita- tively with results obtained by other models, particularly those of Mays and Lloyd (1966), whose work predated ICRP Publication 20 (ICRP, 1972) and those of Papworth and Vennart (1973) All these models indicate than an infant ingesting 37 kBq of 90Sr will receive

a higher dose than an adult ingesting the same amount, and that the doses to the infant may be higher by a factor of as much as eight compared with the adult In fallout studies, a smaller factor of increase, about 2-3, has been observed, but this is to be expected since the higher uptake at an earlier age would tend to be "diluted

or averaged out with lower uptake at later ages

An important feature of the studies of the metabolism of radio- strontium is the measurement of retention This is presented in Section 4 since it bears directly on dosimetry

4 Dosimetry of eOSr

For purposes of this report, the concepts adopted by the ICRP (1977,1979) will be utilized in the main A succinct summary of the dosage formulations, their use, and cautions regarding their use is contained in NCRP Report No 84 (NCRP, 1985) This includes detailed extracts from ICRP documents concerning the dosimetry of bone These will not be repeated here

Essential to all calculations of dose for long-lived radionuclides deposited in tissue is knowledge of the kinetics of retention because the rate of elimination frequently has as much or more influence on dose as does the physical half-life For purposes of this report, it will

be assumed that the retention models, risk estimates, etc apply

at doses and dose-rates sufficiently low to exclude acute radiation effects

Table 34 of Marshall et at., (1973) gives time integrals of the effective retention of 89Sr and 90Sr Part of this table is reproduced here as Table 4.L2 The integral values are mean residence times in

lo4 Bq-days following the introduction of one unit of radiostrontium into the blood whether by injection, by inhalation, or following inges- tion For a n intake to blood of 1 Bq, the time integral of activity will

be in units of Bq-days

The metabolic model of ICRPPublication 20 (ICRP, 1972; Marshall

et al., 1973) enables one to estimate the behavior of following an intake by an adult By far the greatest residence (time integrated activity) is in bone, and the model gives separate estimates for reten- tion in compact and in cancellous bone This is of importance because the active marrow is more closely associated with and would receive most of its dose from the 90Sr + retained in cancellous bone The model predicts (Table 4.1), for 37 kBq of entering the blood of

an adult, a residence of 1474 x lo4 Bq-days for the compact bone

21t will be noted that this Table gives integration times for only one year and 50 years and infinity Frequently, intermediate times may be useful in dose calculations

A table showing microcuriedays accumulated in cancellous and cortical bone and in

soR tissue for 2, 7, 30 60, 180, 365, 1,825, 3,650, 7,300, 10,950, 14,600 and 18,250 days after intake of one microcurie of three strontium isotopes and of barium-140 is given in the Reactor Safety Study, WASH-1400, Appendix VI (NRC, 1975) The data were supplied by Marshall This table, with the data converted to SI units, is reproduced as Appendix "An to this report

TABLE 4.1-Tim integmls of the retention functions in 10" Bq days for 1 year, 50 years and infinity after injection of 3.7 x 10"Bq

including the effect of mdiooctive decay

Time A£ter

"Adapted from Table 34 of Marshall, et al., 1973 For consideration of other times of integration see footnote 2 and Appendix A

8

"l'he total adult skeleton of Reference Man contains 5 kg of mineralized bone, 4 kg as compact bone and 1 kg as cancellous bone $

10 / 4 DOSIMETRY OF SOSr

and 582 x lo4 Bq-days in cancellous bone for the 50-yearperiod post exposure (numerical values taken directly from Marshall, et al.,

1973, and converted to SI units) The residence times corresponding

to infinite residence of 1558 x lo4 and 582 x lo4 Bq-days, respec- tively, are quite close to these values, and consequently, little addi- tional dose would be received even though the person did live beyond the 50-year period post intake In soft tissues, the residences are quite small in comparison to the total mass of such tissues, and because there is no known concentration factor, except possibly for cartilage (Snyder and Cook, 1964), the dose to these tissues can be estimated as an average

The model of Papworth and Vennart (1973) extended the ICRP 20 (Marshall) model (ICRP, 1972) to children to yield separately the retention in compact bone and in cancellous bone Their estimates show the infant to have the largest dose commitment per unit of intake, but still within the range of values obtained by other models Each of the long-term animal experiments, which will be described

in more detail in the next section, provided information on retention

in the skeleton These could be described by a retention function of the form:

where R = percentage retention, k,, k, and k, and A,, A, and A, are determined from experimental data These were used to compute the doses associated with given times and effects that are presented in the Tables of Section 5 Details of the numerical values of the reten- tion functions in the several exponents can be found in Appendix B (derived for this report) Relative magnitudes of corresponding val- ues in Appendix B provide an important comparison of the retention

of administered 90Sr among the various lifespan studies, because data from each laboratory were supplied to a single investigator (NJP) who performed the calculations with a single program This approach should eliminate differences that might arise from differ- ent fitting routines used a t various laboratories or the selection

of animals at differing skeletal dose-rates For animals that were exposed to 90Sr by injection or inhalation, the calculated values of the parameters were roughly similar

In the long-term experiment with monkeys carried out at the University of California, Berkeley, and presented in more detail along with the other long-term experiments in Section 5, the kinetics

of retention were worked out in depth The time involved was over

20 years

4 DOSIMETRY OF 90Sr / 11 Total body retention in this work was described by Durbin et al., (1974) with an exponential equation of 6 terms, viz:

Values of the A's and c's given by Durbin et al (1974) are displayed

in Table 4.2 Correlates of this experiment will be evident in the discussion of dosimetry and derivation of risk

The values for uptake to blood from the GI tract or lung (0.3 with

a range of 0.2 to 0.5) used in ICRP Publication 30 (ICRP, 1979),and other metabolic "constants" contained therein, appear entirely appropriate for adults For subjects exposed at younger ages, the necessary recent modifications of the model for uptake, retention, etc (Papworth and Vennart,l973; Johnson, 1983; Leggett et al.,

1982, 1984; Crawford-Brown, 1984) should be used

For isotopes that emit beta particles of long-range in tissue, such

as 90Sr + it is probably sufficient to use the total residence times for either compact or cancellous bone, respectively, in order to calculate the doses to bone marrow or to endosteal surfaces rather than new or old compact/cancellous bone because the spatial distribu- tion of radiation dose is quite uniform Thus, the columns labeled COMPACT BONE or CANCELLOUS BONE in Table 4.1 (Table 34

of Marshall, et al., 1973) should be used for 90Sr rather than the columns NEW BONE or OLD BONE or BONE SURFACES The rationale for this procedure is that when the particle range is large, the exact location of the activity within a given type of bone should not have an important effect upon the corresponding distribution of dose

In contrast to the long-range beta emitters such as + for the acute intake of radionuclides that emit either alpha particles or beta particles of short range ( 4 0 0 pm), one should take into account the effect of redistribution of the initially deposited material (hotspot burial) This greatly reduces the dose to bone surface or bone marrow from the NEW BONE component of the activity (eg., the method of Marshall, 1962) When intake is by continuous exposure beginning

at very young ages, bone surfaces are labeled continuously, and hotspot burial is not important The diffuse component of OLD BONE

is more significant for isotopes of long half-life (i.e., months to years), and BONE SURFACE is more significant for isotopes of short half- life (i.e., days)

The differences in deposition and retention that occur with time assumed special importance when an ICRP Task Group identified endosteal tissue lying within 10 pm of bone surfaces as the tissue at risk for bone malignancies (ICRP, 1968) Factors, Do, necessary to convert average skeletal doses to the corresponding doses to small

TABLE 4.2-Parameter values for retention in monkeyeb z

aFrom Durbin, et al., 1974

The units of the parameters, "c", are days-'

4 DOSIMETRY OF 90Sr / 13 soft-tissue cavities within bone have been estimated for deriving dose to marrow or to endosteal tissue by Spiers (19681, by Spiers and Whitwell (1972), by Spiers et al (1972), by Whitwell and Spiers (1976) and by Beddoe and Spiers (1979) It is recommended that ratios of the endosteal dose or marrow dose to average skeletal dose,

Do, be used if such a calculation is to be made Dose to other specific components of the skeleton, such as the exterior surface, can be calculated by means of an appropriate ratio (see Figure 1 of Beddoe and Spiers, 1979) However, data of Lloyd and Henning (1983) indi- cate that the location of the cells at risk for bone sarcoma induction cannot be identified as a single layer within 1-10 km of bone sur- faces This suggests that it may be premature to calculate local dose

to cells at risk until their identity and location can be defined more adequately Furthermore, there is not yet general agreement as to the proper way to determine dose to the critical cells for bone sarcoma induction (Jee, 1984) Fortunately, the problem is trivial for gOSr + because of the range of the beta particles emitted So long

as risk coefficients to be compared are expressed in consistent units, e.g., bone sarcomas/104 Gy average skeletal dose, these nuances of dose parameters make little difference for strontium-90

We can turn briefly to the mechanics of a dose calculation for radiostrontium The mineralized bone of reference man has a mass

of 5000 g, and the masses of compact and cancellous bone are taken

as 4000 g and 1000 g, respectively (ICRP, 1975) The dose, D, to these portions of bone (assuming complete absorption of the beta energy)

is given by

where 1.38 x 10" represents (86400 dday) (1000 glkg) (1.602 x 10-l3 J/MeV), 1.075 is a factor representing the greater dose within a small soft tissue cavity compared with the dose to surrounding bone (Spiers, 1968) and can be included or not depending on which dose

is to be estimated, U represents Bq-days per gram of bone and E is the average energy released in Mev per disintegration of 90Sr Because has a decay half-time of only about 2.7 days, it is assumed to be in equilibrium with its parent While this assumption may not be strictly accurate, it is conservative and is not a significant factor in the case of bone (Arnold et al., 1955) There is some evidence that the dose to active bone marrow and to endosteal cells per unit concentration (Bq gOSr/g Ca or Bq/g of tissue) should vary with age

It may be higher for the newborn than for the adult human by about 25% because the marrow spaces are smaller in infants and children The factor becomes less as the marrow spaces approach adult size (Spiers and Whitwell, 1972; Beddoe and Spiers, 1979) If the marrow

14 / 4 DOSIMETRY OFSOSr

dose rate is about half that to the skeleton (Mays and Lloyd, 1972~1, the active marrow would receive a dose commitment of about 1.6 pGyIBq of intake for the newborn, and most of the resulting dose commitment would be received within the first year post intake This would be the dose commitment the infant would receive only if the 90Sr were ingested during the first month or so of life In succeeding months the dose commitment per unit intake would decrease rather rapidly, being about a third of the value for intake a t 0-1 month of age if the is taken a t 1 year of age The average dose commitment

to endosteal cells (an average over the distance 1-10 pm of the bone surface) would be approximately twice that received by the bone marrow

A factor of 0.2 was used in the ICRP model to represent the absorp- tion from the gastrointestinal tract to blood (Table 27, ICRP, 1972) This value is consistent with the data used in developing the reten- tion model as well as with data on retention in fallout even though it differs from the one used in ICRP Publication 30 (viz: 0.3, ICRP 1979)

We will consider the skeletal dosimetry of in the monkey The calculated average skeletal dose-rate, dD/dt, in an adult monkey having 150 g of skeleton per kg body is

=

lo-' Bq MeV day 15Oglk.g

= 1.04 x 10-I IRF Gylday

for an average beta particle energy of 1.13 MeV per disintegration (dis) and where F is the fractional absorption of 90Sr + 90Y beta- particle energy by the skeleton (0.7, 0.75 or 0.8) for a typical 3.3 kg juvenile, 4.6 kg adolescent or 7 kg adult monkey, (Parmley et al., 1962), I is the intake and R is the fractional retention in the skeleton

Cumulative average skeletal dose D can be calculated as

For a n adult monkey living 12 years after the intake of 1 MBq 90Sr, D is about 36 Gy

of California, Davis (UCD), beagles exposed by inhalation at the Lovelace Inhalation Toxicology Research Institute (ITRI), and a review of long-term experiments with 90Sr in monkeys as studied at the University of California, Berkeley, and briefly at the University

of Rochester

In each case, the summary tables and descriptive text represent

an enormous consolidation and extraction of the essence of what were large, long, difficult and expensive experiments Glimpses of the true scope of the efforts can be seen in the occasional reports and publications referenced as appr~priate.~

5.1 Studies with Mice at Argonne National Laboratory

During World War 11, long-term studies on strontium were carried

out with mice at the Metallurgical Laboratory of the Manhattan District These were published post-war from the Argonne National

3Recently the proceedings of a major symposium entitled "Life-Span Radiation Effects Studies in Animals: What Can They Tell Us?" held a t Richland, Washington,

in October 1983 have been published (Thompson and Mahaffey, 1986) This includes individual reviews of the life-span studies with strontium a t the University of Califor- nia at Davis, at the Lovelace Inhalation Toxicology Research Institute (fTRI), and at the Radiobiology Laboratory, University of Utah These reviews serve to update a portion of the information used in depth for the preparation of this report However, the publication of the symposium proceedings came aRer completion of the major effort for this report These later summaries did not result in major revision of the conclusions already drawn, since the bulk of the data had already been gathered and considered

Laboratory by M P Finkel and her colleagues The results were useful, even during World War 11, via the use of in-house reports, for the establishment of the early maximum permissible exposures to radiostrontium

The results have been summarized many times in the open litera- ture and provided the original input for the toxicity ratio approach

to setting the maximum permissible burdens (and derived figures) for many bone-seeking elements (A toxicity ratio is the ratio of doses from two different radionuclides that produce the same level of effect

in the same kind of animal model.) The reader is referred to publica- tions by M P Finkel (1956), Finkel and Biskis (1959 and 1968) and

a review by Mays and Finkel (1980) These data will be used in this report largely for comparison purposes

5.2 Beagles Injected at the University of Utah

This 90Sr experiment was conceived in the early 1950s, it was begun in 1955, animals were injected until 1966, and the study is just now reaching final data analysis

The total of 87 young adult dogs were each given a single intrave- nous injection of 90Sr at dose-levels ranging from 2.1 x lo4 to 362 x

lo4 Bqkg and retained for life-span observation A companion study using single injections of 0.027 x lo4 to 38.5 x lo4 Bqlkg of 226Ra in

120 young adult beagles was also conducted at the same lab~ratory.~ There were 125 control beagles including 57 that were sham injected during the same time interval as the dogs given 226Ra and 90Sr Veterinary care, preparation of the injection solutions, injection pro- cedure, animal selection, experimental design, radionuclide reten- tion and dosimetry have been described in earlier publications (Dou-

gherty et al., 1962; Lloyd et al., 1976) All of the animals given 90Sr

have now died, mainly from causes unrelated to radiation, particu- larly at levels below 121 x lo4 Bqlkg In addition, all of the young adult dogs given 226Ra have died The most prominent goSr-induced endpoint was bone sarcoma Neoplasia of the soft tissues near bone

in the oronasopharynx and paranasal sinuses, and bone marrow dysplasia were observed in excess of the incidence in the control population (P<0.05) but in lower incidences than bone sarcoma

T h e dosage levels were chosen in relation to the bench mark level of the maximum permissible body burden of radium in man The key level (Level 1) for *Ra was equivalent to ten times the maximum permissible body content Other levels (8 total) extended fiwm 119 to 162 times Level 1

5.3 BEAGLES FED g9Sr AT THE IABORATORY 1 17 Leukemia was not observed, and the incidence of myelolymphoproli- ferative disease was not elevated Analysis of the dose-response rela- tionships for both 90Sr and 226Ra for bone sarcoma induction have been reported (Mays and Finkel, 1980) Table 5.1 summarizes the results of these experiments in updated and consolidated form Inci- dence of bone sarcoma as a function of calculated radiation dose is given in Figure 5.1 for both 90Sr and 226Ra in the Utah studies, omitting, for clarity, the data points beyond the region where the effect is maximized for 226Ra (100%) and the lowest three zero inci- dence points for 90Sr

The studies of Y3r and 226Ra effects at the University of Utah can

be compared with corresponding experiments by Miriam Finkel and her colleagues at Argonne National Laboratory, already mentioned , (Section 5.1) (Mays and Finkel, 1980).5 Both of these studies indicate that the effectiveness (toxicity) ratio of 226Ra relative to 90Sr (1) may approach unity at high doses, (2) increases markedly as the skeletal dose is reduced, (3) approaches a value of about 25 for the lowest doses at which definite effects from 90Sr were seen (see Table 5.2 and Figure 5.2), and (4) may be extrapolated to even more than 25 at extremely low levels of skeletal dose and dose-rate such as might be expected in the human population from normal peacetime environ- mental exposure to 90Sr Results of more recent experiments reported

by Reif and Triest (1982) with 90Sr in C57BLIM female mice indicate that for injection levels between 118 x lo4 and 1180 x lo4 BqJkg, tumor incidence is roughly comparable to that found at correspond- ing levels for the female CF1 mice at Argonne (Mays and Finkel, 1980) This analysis provides clear evidence that the strontium + yttrium-90 beta particles are much more effective per Gy at high doses and dose rates than at low doses and low dose rates

The time from injection to death with osteosarcoma ranged between about 2.6 and 13.3 years There were no other major late effects induced by wSr that were as important as bone malignancy For the two groups with the lowest dose-rate at which a bone sarcoma was seen, the survival of the animals with osteosarcoma was similar

to the respective mean

5.3 Beagles fed at the Laboratory for Energy Related Health Research (LEHR), University of California at Davis

A total of 389 beagles at the University of California, Davis (LEHR), were exposed to 90Sr by feeding from the onset of fetal

%is is a retmspective analysis of the most cogent data from this large series of experiments and cites many of its original papers

TABLE 5.1-Bone sarcomas in beagles at the University of Utah given 90Sr or ZZ6Raa

Av Skel

aAdapted from Mays and Finkel, 1980, but with revised dosimetry (Wrenn, 1984) updated, 1987

bIncluding one recently diagnosed ameloblastoma (skeletal malignancy)

There were also observed among the 2BFki-injected dogs 1 fatal case of blood dysplasia in the 38.5 x 104Bq/kg group (at 1288 days), 1 case

of fatal malignant tumor of the tympanic bulla in the 1.25 x 104Bq/kg group (at 4615 days) and 3 fatal cases of soft tissue neoplasia (non-

melanoma) of the oral cavity (0.164 x 104,0.229 x lo4 and 0.081 x 104Bq/kg groups a t 3254,4003 and 3848 days respectively) Among the

goSr-injected dogs, there were 5 animals that died with blood dysplasia (2 a t 362 x 104Bq/kg a t 990 and 1021 days, 1 each a t 235 x lo4, 121

x 104 and 40 x 104BqIkg a t 1493,2114, and 1285 days), 4 with malignant tumors of soft-tissue in the nasopharynx (1 a t 362 x lo4, 2 a t

235 x 104 and 1 a t 121 x 104Bqkg a t 1982,2066,2813 and 4226 days, respectively) and 4 with soft tissue neoplasia (non-melanoma) of the

oral cavity (2 at 121 x lo4, 1 each a t 40 x lo4 and 2.11 x 104Bq/kg a t 2093,4844,2898 and 5363 days after injection, respectively) Among

125 control dogs in the entire colony, there was one fatal case of blood dysplasia (leukemia, 3971 days), and 1 case of soft-tissue malignancy

of the nasopharynx (5203 days) According to Fisher's exact Test for Independence (Sokal and Rohlf, 1969), there was no significant difference

in occurrence of these soft-tissue malignancies and blood dysplasias between the control animals and the dogs given ZZ6Fia, but there was a

clear excess ( P < 0.05) of blood dysplasias and of soft tissue malignancies (non-melanoma) of the oronasopharynx in the dogs given goSr a s

compared to the control group For all other soft-tissue tumors, the difference between the occurrence among the %r or 226Ra groups and the

group of 125 control dogs was not significant (P > 0.05) except for introcular melanoma in dogs given Ra

dTumor was discovered in 1987 re-analysis of slides

5.3 BEAGLES FED SOSr AT THE IABORATORY / 19

BONE SARI

Fig 5.1 Bone sarcoma incidence as a function of average skeletal dose from 90Sr

(open squares) or 226Ra (closed circles) This updates to 1987 the data contained in Table 5.1 No bone sarcomas were induced by 90Sr in dose groups below about 20 Gy, while bone sarcomas were induced by n6Ra in a group averaging 0.85 Gy

TABLE 5.2-Relative effkctiveness for bone sarcoma induction of 226Ra and 90Sr in

beagles and mice as a function of avemge skeletal dose

'From Mays and Finkel (1980)

bValues interpolated from updated Figures 1 and 2 of Mays and Finkel (1980)

20 1 5 EFFECTS OFsoSr

0

AVERAGE SKELETAL DOSE I N GY 100 DAYS BEFORE TUMOR

APPEARANCE OR 140 DAYS BEFORE DEATH

Fig 5.2 Bone sarcoma incidence as a function of average skeletal dose from %r (open squares) or 22BRa (solid circles) in female mice studied at Argonne National Laboratory (Mays and Finkel, 1980) The shape of the dose response curve for %r in mice resembles that for 90Sr in beagles (Figure 5.1) and is much different from the curve for 22%a in mice or beagles

calcification in utem to 540 days of age a t seven levels ranging from 0.026 x lo4 to 37 x lo4 Bqtgram of dietary calcium (Table 5.3a) The 90Sr was uniformly mixed into the diet fed each day There were also 80 control beagles that were introduced into the experiment during the same period An additional 46 beagles were each given a single intravenous injection of 13.7 x lo4 or 122 x lo4 Bq goSr/kg

a t 540 days of age (Table 5.3a) In a companion experiment, 251 beagles were each given a series of 8 semi-monthly injections of 226Ra beginning a t 435 days of age (Table 5.3b) Six dosage levels were employed, ranging from 0.089 x lo4 to 37 x lo4 Bqkg There were also 84 controls introduced during the same period Experimental design, preparation of the injection solutions, choice of dosage levels, veterinary care, animal selection, radionuclide retention and dosim- etry have been described previously (Book, 1980; Raabe et al., 1981a)

The tables give the incidence of effects in both bone and soft tissue Hematopoietic effects, including myeloproliferative disorders (Dungworth, et al., 1969), were seen mainly at the three or four

highest levels of 90Sr ingestion (average skeletal doses of about 20

5.4 BEAGLES EXPOSED TO AIRBORNE 80Sr / 21

Gy and above, but with one observed a t 6.28 Gy and two in control dogs 15.4 and 16.4 years old) and one a t 6.24 Gy among dogs injected with Y3r At later times post-exposure, bone sarcoma, head sinus carcinoma and squamous cell carcinoma of the tissues in the mouth appeared at lower levels of exposure in clear excess of their occur-

rence in the control dogs (Book et al., 1981, Pool, et al., 1972 updated

by Raabe 1990) Similarly, carcinomas of soft tissues overlying bone

in 90Sr contaminated rabbits were reported by Vaughan and William- son (1969)

Comparison of bone sarcoma incidence vs dose in beagles given 90Sr or 226Ra (Tables 5.3a and 5.3b) indicate that their relative effec- tiveness (toxicity ratio) is similar at high doses, but at lower levels, the relative effectiveness of 90Sr is decreased, reaching a value of about 1/12 that of radium at the lowest level at which wSr induced

tumors have appeared (Raabe et al., 1981a)

5.4 Beagles Exposed to Airborne 90Sr at Inhalation

Toxicology Research Institute (ITRI)

Studies of effects in beagles exposed by inhalation (McClellan

et al., 1973) have been completed a t the Lovelace Inhalation Toxicol- ogy Research Institute (ITRI) in Albuquerque, NM (Tables 5.4a and 5.4b) Sixty six adult beagles of both sexes were exposed to Y3rC1,

by inhalation, resulting in initial body burdens ranging from 3.59

x lo4 to 703 x lo4 Bqlkg All of these dogs are now dead Animal selection, veterinary care, experimental design, aerosol preparation, method of exposure, radionuclide retention and dosimetry have been

reported (McClellan, et al., 1972) Tissue distribution, resulting radi-

ation doses, and biological effects in the skeleton have been found to

be similar in dogs exposed by inhalation to those in dogs exposed by injection

The 90Sr inhaled as the chloride was rapidly translocated from lung to skeleton (McClellan, et al., 1972) It appears that once 90Sr

as the chloride reaches the blood it is handled as though it were given intravenously Thus, the authors reported the observed biological effects in terms of the long-term retained burden of 90Sr

The most prominent early effect was a dose-related pancytopenia

(Gillett, et al., 1987a) The major finding among the long-term survi- vors was an excess of bone tumors (Gillett, et al, 198713) Most of the

dogs with long-term retained burdens of 1.0 MBqkg or greater died with a bone tumor There were three nasal tissue tumors, and two dogs died of myelomonocytic leukemia (for more detail see McClel-

lan, et al., 1983)

TABLE 5.3a-Beagles given 90Sr at LEHR Data are as of January 1,1988 (no dogs alive) Shown are the number of dogs with bone sarcoma, myelolymphoprolifemtive syndrome (MPS), and malignant tumors of soft tissues near bone (mainly squamous cell carcinomas of

tissues in the mouth) for dogs dying or terminated from the experiment

TABLE 5.4a-Summary of individual body burdens and T+related cause of death in 66 dogs that inhaled graded activity levels of 90SrC1,

a n d were maintained for life-span observation (ITRZ)"

Days post exposure

Bone marrow hypoplasia m C3

Bone marrow hypoplasia Bone marrow hypoplasia 8

Bone tumor Bone marrow hypoplasia Bone tumor

Bone tumor Bone tumor Bone tumor Bone tumor Bone tumor

-

Bone tumor Bone tumor Bone tumor Bone marrow hypoplasia Bone tumor

Bone tumor Bone tumor Bone tumor Bone tumor

TABLE 5.4a-Continued Q,

Days post exposure

"From Gillett, et al., 1987b, Inhalation Toxicology Research Institute, Albuquerque, NM 8

'D = Died; E = euthanized

dImrnediate cause of death was pyometra

TABLE 5.4b-Summary of Vr-exposed and control dogs dying of causes unrelated to the inhalation of SCSr Davs after

exposure Age at Primary cause of death

Dog No Sex to death death days LTRB,MBqlkg Neoplastic Nonneoplaatic goSr-exposed dogs, neoplastic cause of death

goSr-exposed dogs, nonneoplastic cause of death

Epileptic seizures Hemorrhage, cerebellar Accidental death Peritonitis Malabsorption syndrome Hepatitis

Congestive heart failure Hepatic degeneration Nephrosclerosis Congestive heart failure Ne~hrosclerosis Osteoarthritis, spine and stifles Ne~hrosclerosis

TABLE 5.4b-Continued h3 00

Days after

Control dogs, neoplastic cause of death

P 9E F 2,638 3,037 0 Fibrosarcoma, thoracic wall

40D F 3,654 4,041 0 Adenocarcinoma, mammary gland

24E F 4,508 4,901 0 Carcinoma, thyroid gland

158A M 5,008 5,446 0 Squamous cell carcinoma, tonsil

13D F 5,057 5,440 0 Adenocarcinoma, mammary gland

33B M 5,103 5,491 0 Lymphosarcoma

160A M 5,482 5,919 0 Squamous cell carcinoma, tonsil

5.5 EFFECTS OF 90Sr FED TO MINIATURE SWINE 1 29

Although a few animals a t the highest dose levels died with hema- tological disorders, bone neoplasms were the most prominent radia- tion-induced endpoints at lower levels, occurring in 17 of the 48

lifespan dogs at between 1 and 10.36 MBqkg initial body burden, but in just one dog below 1 MBqkg There were 13 animals with bone sarcoma among 24 other dogs at 1.59 to 5.55 MBqIkg initial body burden in a sacrifice study Only two leukemias (myelomono- cytic) were observed There were also a total of 3 malignancies of tissues in the oral cavity and nasopharynx

There were also experiments at ITRI in which 90Sr was inhaled in insoluble form No bone tumors occurred in these animals, although some neoplasms did occur in lung This work is not included in this summary since it does not contribute to the objective of deriving risk estimates for bone cancer and leukemia Details are contained in ITRI annual reports and publications

5.5 Effects of 90Sr Fed to Miniature Swine at Hanford

at 9 months of age, but those in the second and third generations

were exposed to 90Sr beginning in utero Sr was given each day in a

"spiked feed pellet Following birth the animals were exposed to

*Sr via the dam's milk (McClellan and Bustad, 1963; McClellan et

al., 1962; McClellan, 1964) At six weeks of age, the animals were weaned and began receiving *Sr daily in their diet at one-fourth that of their dam At three months of age their dietary intake was increased to one-half the 90Sr of their dam's diet and at six months

to an amount equal to that of their dam

Information bearing on the retention and dosimetry of 90Sr in these

swine has been published (Clarke et al., 1969; Palmer et al., 1970;

Ragan et al., 1972; Mays and Lloyd, 1972b; Krusemark et al., 1974;

McClellan, et al., 1967; McClellan, 1965)

Hematopoietic effects from irradiation of bone marrow including neutropenia, lymphopenia, thrombocytopenia, and myeloprolifera- tive disorders with myeloid and histiocyte infiltration of tissues such

as kidney, heart, testes, and lung were observed among animals at

the highest feeding levels (Tables 5.5a and 5.5b), but these were not seen in excess of the control incidence in pigs at low doses and dose

rates (Howard and Clarke, 1970, McClellan, et al., 1963a; McClellan,

1966) Dental lesions, uterine pathology and arthritis were among the most common findings at death in both irradiated and control

animals (Ragan et al., 1972) Farrowing performance (litter size, stillbirth and birth weight) was not affected at intake levels of

TABLE 5.5a-Summary of $OSr effects in miniature swine Studied at Battelle Pacific

Northwest Laboratories (Richland, WA) Feeding Level

Pancy topenia-Hemorrhagic 3 months crisis, Myeloid Metaplasia

TABLE 5.5b-Neoplasia in female miniature swine ingesting daily (combined fi

and f2 generation, exposure began in utero) Studied a t Battelle PNL

T w o of these tumors were malignant

'All ovarian tumors were malignant

dFour of these tumors were malignant

eSeven of these tumors were malignant

'Four of these tumors were malignant

NO malignancies

hSix of these tumors were malignant

5.6 Strontium-90 Exposure Via Injection in Monkeys 1 31

23 MBqIday and lower It was concluded that feeding levels of wSr high enough to affect fetal or neonatal mortality in this species will

not permit survival of the dam through gestation (Clarke et al., 1970b; McClellan, et al 1963b)

Bone tumors, classified as osteosarcomas or giant cell tumors, were found in the skeletons of many (see Table 5.5b) but not all of the 90Sr treated animals Most of these tumors occurred in the skull, including the mandible and maxilla All of the giant cell tumors

were found in the mandible (Howard et al., 1969) A comparative

study of the histopathologic effects in bone of injected 226Ra and @OSr

in the miniature pig (Clarke, 1962) verified that 226Ra was more damaging than 90Sr as measured by incidence of bone tumors at low doses There appeared to be an increase in general sofi-tissue cancer incidence among animals exposed to dietary 90Sr; this was attributed

in part to blood-borne @OSr (Palmer et al., 1970) Note that this

includes increases in liver neoplasia at the moderate doses, but only two liver tumors were malignant

A summary of the primary effects and mean survival times in miniature swine at the several dosage levels is given in Table 5.5a Additional details are given in Table 5.5b

5.6 Strontium-SO Exposure Via Injection in Monkeys (University California, Berkeley, Supplemental Data from

the University of Rochester) Forty juvenile, adolescent and adult male and female Rhesus mon- keys aged 2 to 12 years and weighing 2.6 to 9.4 kg were given 0.13 to 6.21 MBq by single injection (0.018 to 0.14 MBq goSr/kg)

In addition, 2 short-term animals were injected with 1.85 MBqIkg Retention was determined by excreta collection and by periodic in

rence Berkeley Laboratory (LBL), University of California at Berke- ley, for up to about 20 years after entry into the experiment The most pertinent metabolic parameters for radiostrontium in these animals are shown in Table 4.2

No biological effects attributable to the 90Sr exposure could be detected in this experiment, even over a twenty-year period Never- theless, the data currently available must be considered as still preliminary The study can provide a useful upper limit on risk for bone sarcoma and leukemia in a primate species (see Section 8.1, footnote 8, for an example of this method)

In another experiment (Casarett et al., 1962) that was done at the

University of Rochester, 7 adult monkeys were given 1.85 or 3.7 MBq

'"Sr orally (by gavage, single dosage) There were 2 cases of bone sarcoma, occurring a t 36 months (chondrosarcoma) and 45 months (osteosarcoma) after administration Skeletal dose for the animal dying with osteosarcoma a t 45 months has been calculated (Mays and Lloyd, 1972b) as 34 Gy a t death and 25 Gy to the estimated start

of tumor growth

Details of life span and cause of death, where known, are embodied

in Appendix C for the monkeys studied a t LBL Few, if any, of the causes could be reasonably associated with radiation exposure

Experiments With 90Sr

The animal studies have shown that although the effects can occur with relatively high incidence a t high levels of radiation dose from 90Sr they occur a t only a very low incidence, or not a t all, a t the lower

levels of radiation dose from this deposited radionuclide (i.e., no bone

sarcomas were seen a t individual average skeletal doses in the Utah beagle study between about 1 and 18 Gy) This is in contrast to the effects of radium (see next section) Clearly these "lower levels" of dose are well above those to be expected in humans from all normal peace-time operations No striking differences from previous analy- ses of these experiments have been seen in this analysis, and it is unlikely that results from the few animals still alive (at the time this report was prepared) will alter the basic conclusions reached High doses to skeletal tissues from 89Sr or 90Sr have produced bone sarcomas, carcinomas of the nasopharynx and head sinuses, squamous cell carcinomas in tissues within the mouth, or hematopoi- etic neoplasia (leukemia) and dysplasia in the following experimen- tal animals: mice (Finkel and Biskis, 1959; Nilsson, 1972; Ito et al., 1969; Van Putten and DeVries, 1962; Brooks et al., 1974); rats (Moskalev et al., 1969; Casarett et al., 1962; Jones et al., 1970);

cats (Ward et al., 1972, Nelson et al., 1973); rabbits (Vaughan and Williamson, 1969); monkeys (Casarett et al., 1962; Durbin et al., 1974); sheep (McClellan and Jones, 1969); dogs (Goldman et al., 1969, 1972; McClellan et al., 1972; Finkel et al., 1972; Litvinov, 1962; Dougherty et al., 1972; Mays et al., 196913; Raabe et al., 1981b; Gillett

et al., 1987a, 1987b); and pigs (Clarke et al., 1972; McClellan, 1964)

There is clear evidence in the preceeding references that the dosage pattern as well as the total dose can have a significant effect on the outcome of exposure The same applies to the age a t irradiation For example, hematopoietic effects and soft-tissue malignancies (squa-

5.7 Summary and Discussion 1 33 mous cell carcinoma) were important endpoints in dogs fed beginning in utero but much less significant among adult dogs that inhaled or were injected with 90Sr even at comparable skeletal doses

No effects attributable to wSr have been observed in the monkeys

a t the Berkeley project presumably because of the low doses given However, bone sarcomas have been induced by Y3r in other monkeys given much higher dosages (animals from University of Rochester transferred to Berkeley, Casarett et al., 1962) Myeloproliferative

disorders have been observed in both dogs and swine fed 90Sr begin- ning as young animals, but no clear excess of these above control incidence has been detected in dogs that inhaled or were injected with as adults Bone sarcomas have been induced by 90Sr given

by injection or in feed to swine and in dogs given 90Sr by injection,

by inhalation, or in feed Malignancies of soft tissue near bone appar- ently have been induced in dogs given 90Sr by injection and through the diet Bone sarcoma induction and hematopoietic effects were also

observed among cats given 89Sr (Ward et al., 1972; Nelson et al.,

1973) Because of these observations there is reason to expect similar effects (myeloproliferative disorders, bone sarcomas, cancer of soft tissue near bone) in humans receiving sufficient doses from +

wY or 89Sr

Only at the highest levels of exposure has there been a signifi- cant effect upon lifespan and, as already noted, fertility and fecundity seem to be unaffected below potentially lethal doses This also con- trasts markedly with the effects of radium on lifespan and bone sarcoma induction

These data can be used to estimate the risk to humans from incor- porated 90Sr when appropriate results from animal experiments with 226Ra and experience with 226Ra in humans are included in the analy- sis This is done in the following section

The major goSr-induced effects in these animal experiments were observed at times after exposure that were generally quite long at dose levels near the range of projected human exposures from normal peacetime operations The time course of the appearance of these effects a t low doses is less important than if gOSr-induced effects had appeared throughout the post exposure period

6 Effects of 90Sr Compared

A comparison of bone sarcoma induction by the bone-seeking radio- nuclides, 226Ra and 90Sr, is important because a considerable body of data is available on exposure of man to 226Ra; this includes informa- tion on more than two thousand cases of exposure to radium at a wide variety of levels Moreover, these cases have been followed for many years, and extensive studies of the sequelae and of their frequency have been reported (Evans, 1966; Rowland et al., 1971; Rowland and Stehney, 1977; Gustafsen and Stehney, 1983) These studies with radium served as the principal support for the widely used limiting occupational body burden of 3.7 kBq 226Ra which was recommended by the NCRP in 1941 (NCRP, 1941) Many of the extensive, long-term animal studies mentioned earlier in this report were based on the idea that a comparison of relative effectiveness of 90Sr and 226Ra in a variety of species would offer a basis for assess- ment of the hazard to man from 90Sr Generally, such comparisons were made by computing ratios of doses at which a common incidence

of effect was seen in each of the various species studied

Both the dosimetry and biological effects of 90Sr and 226Ra found from the study of mice at Argonne National Laboratory (ANL), Illi- nois, have been analyzed by Finkel and Biskis (19681, Marinelli (19691, Mays and Lloyd, (1972b) and Mays and Finkel (1980) in terms of these "toxicity ratios." They indicate that for an equal percent of bone sarcoma occurrence, 226Ra is from 1 to 25 times as effective as 90Sr on the basis of average skeletal dose (Table 5.2); the relative effectiveness is a function of dose The studies of Finkel and Biskis (1968) in mice indicated that at least 200 times as much activity of 90Sr as of 22%a must be injected to obtain about a 10 percent incidence of bone sarcomas Analysis by Marinelli (1969) of the dosimetry applicable to the work of Finkel and Biskis (1968) indicated that the ratio of absorbed doses averaged over the skeleton

is 8.8 or more (Marinelli, 1969) In the case of these mice, strontium and radium have essentially the same retention function, so the ratio

of average skeletal doses is essentially independent of time Below the 10 percent level of incidence, statistical difficulties obscure the interpretation of the data